R Notebook

library(tidyverse)
library(cluster)
library(factoextra)
library(dendextend)
library(broom)
library(animation)

computers

NA

computers %>%
  ggplot(aes(x = hd, y = ram)) +
  geom_point()

in answer to the suitability to cluster…. I dunno, kinda…..Logic is telling me yes otherwise what’s the point in the hiomework but looking at the data it looks a bit organised and non linear. So….who knows?? Yes i suppose it is, there’s visually soe grouping on the y axis (RAM) around like 4, 8, 16 (GB?) with a wider spread on the x axis.

computers_clean <- computers %>%
  select(c(ram, hd))

computers_clean

summary(computers_clean)

      ram               hd        
 Min.   : 2.000   Min.   :  80.0  
 1st Qu.: 4.000   1st Qu.: 214.0  
 Median : 8.000   Median : 340.0  
 Mean   : 8.287   Mean   : 416.6  
 3rd Qu.: 8.000   3rd Qu.: 528.0  
 Max.   :32.000   Max.   :2100.0  

First i’m gonna scale the data.

computers_scale <- computers_clean %>%
  mutate_all(scale)

computers_scale

clustered_computers <- kmeans(computers_scale, centers =  4, nstart = 25)

clustered_computers

K-means clustering with 4 clusters of sizes 419, 904, 2235, 2701

Cluster means:
         ram         hd
1  2.4276068  2.5442566
2  1.3728674  0.5699667
3 -0.1367705  0.2706146
4 -0.7229016 -0.8093732

Clustering vector:
   [1] 4 4 4 4 2 2 4 4 4 4 4 4 4 4 4 4 4 4 4 3 4 4 2 4 3 4 4 3 2 4 4 4 4 4 4 4 4 3 4 4 4 3 4 4
  [45] 2 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 3 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4 4 4
  [89] 4 3 4 4 4 4 2 4 4 4 2 4 4 4 4 4 4 4 4 4 4 4 3 4 3 2 4 4 4 4 4 4 4 4 4 3 4 4 4 4 4 4 4 4
 [133] 4 4 4 4 4 4 4 4 4 3 4 4 3 4 4 4 4 3 4 4 3 4 4 4 4 4 4 4 4 4 4 4 2 4 4 4 4 4 4 4 4 4 3 4
 [177] 4 4 4 4 4 4 4 4 3 4 4 4 2 2 4 4 4 4 4 4 4 4 4 3 3 4 4 4 4 4 4 3 4 3 4 4 4 4 4 4 4 4 2 4
 [221] 4 4 4 3 4 4 4 2 4 4 4 4 4 4 4 4 4 4 4 3 4 4 4 4 4 4 2 4 4 4 4 4 4 4 4 4 4 4 4 2 3 4 4 4
 [265] 4 4 4 4 2 4 4 3 4 2 4 2 4 4 4 4 4 4 4 4 4 4 4 4 2 4 3 4 4 4 4 2 4 2 4 4 4 4 4 2 4 4 4 4
 [309] 4 3 4 4 2 3 4 4 4 4 4 4 4 2 2 4 4 4 4 4 2 4 4 4 4 4 4 2 4 2 4 4 4 4 3 4 4 4 4 4 4 4 4 4
 [353] 4 4 4 4 2 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 2 2 4 4 4 4 2 4 4 4 4 4 4 4 4 4 4 4 4 2
 [397] 4 4 4 3 3 4 3 2 2 4 4 4 3 2 4 4 4 4 2 4 4 4 3 4 2 2 4 4 4 4 2 3 4 3 2 2 2 4 3 4 4 4 4 2
 [441] 4 4 4 4 3 2 4 4 4 4 4 4 4 4 4 4 4 4 4 2 4 4 3 2 4 4 3 4 3 3 4 4 4 3 4 4 4 4 4 4 4 4 4 4
 [485] 4 3 4 3 4 3 1 3 4 3 3 4 2 4 4 4 4 4 4 3 4 4 3 2 3 3 3 4 2 4 4 4 4 4 4 4 4 3 4 4 4 2 3 4
 [529] 4 4 3 4 2 4 4 4 4 4 3 4 2 4 4 4 3 3 4 3 4 4 3 4 2 4 2 4 4 2 2 4 2 4 4 4 4 4 4 4 4 3 4 4
 [573] 4 4 4 4 4 2 3 4 4 3 3 4 3 4 2 4 4 4 4 4 2 4 4 4 3 3 4 4 4 2 4 2 4 4 4 4 4 4 2 3 3 3 4 1
 [617] 4 4 4 4 4 4 3 4 4 3 3 4 4 4 4 4 4 4 4 4 4 4 4 3 4 4 4 2 4 2 2 3 4 4 3 3 3 3 4 4 4 4 4 4
 [661] 2 4 4 4 4 3 2 4 4 3 4 2 3 2 4 3 4 4 3 4 4 4 3 3 4 3 3 3 4 3 4 4 4 2 4 4 3 4 2 4 4 4 4 2
 [705] 4 4 2 4 4 4 2 4 4 3 4 4 2 2 3 3 1 4 4 4 4 4 4 2 3 4 4 4 4 4 4 4 4 4 2 2 2 4 4 4 4 2 3 4
 [749] 4 3 4 3 4 2 4 4 4 4 4 4 4 4 4 2 4 4 2 4 4 3 3 4 2 2 2 3 4 4 2 3 4 4 2 3 4 4 4 3 4 2 4 2
 [793] 4 3 4 3 2 4 2 3 4 2 4 3 4 3 4 3 4 4 3 4 4 2 4 4 3 4 4 4 3 3 4 3 2 4 4 4 4 2 3 4 4 4 4 4
 [837] 2 3 4 4 3 4 3 2 3 4 4 4 4 4 4 3 3 4 2 3 4 4 4 3 3 4 4 4 3 4 4 3 3 3 4 4 3 4 4 3 4 4 4 4
 [881] 4 2 4 3 2 4 4 2 4 4 3 2 4 4 3 4 2 2 4 3 1 3 2 4 3 4 4 4 2 4 2 4 4 4 4 4 4 4 4 4 3 4 3 4
 [925] 4 4 4 4 3 4 4 2 4 4 4 3 3 3 4 4 4 2 4 4 3 3 4 3 4 4 2 3 4 4 2 3 4 3 4 2 3 4 3 4 3 2 4 4
 [969] 4 3 3 3 3 4 4 3 4 4 4 4 4 1 2 4 3 3 4 4 4 4 4 4 4 1 2 3 3 4 3 4
 [ reached getOption("max.print") -- omitted 5259 entries ]

Within cluster sum of squares by cluster:
[1] 452.8436 286.4189 787.9572 423.3334
 (between_SS / total_SS =  84.4 %)

Available components:

[1] "cluster"      "centers"      "totss"        "withinss"     "tot.withinss" "betweenss"   
[7] "size"         "iter"         "ifault"      

Next i’m going to use

Tidy

Glance

Augments

to have a look at the clusters and the data contained within in a bit more detail.

Here i’m looking at info about the mean and the size of each cluster.

tidy(clustered_computers, 
     col.names = colnames(computers_scale))

Now i’m looking at info about the full clustering.

glance(clustered_computers)

This function shows me which cluster each data point has been added to. (comparing the new data i’ve clustered with the old data. )

augment(clustered_computers, computers)

computers_scale %>%
  kmeans.ani(centers = 4)

Now i’m going to look at how many clusters i need in order to have the max heterogenity.

max_k <- 20

k_clusters <- tibble(k = 1:max_k) %>%
  mutate(
    kclust = map(k, ~ kmeans(computers_scale, .x, nstart = 25)),
    tidied = map(kclust, tidy),
    glanced = map(kclust, glance),
    augmented = map(kclust, augment, computers)
  )

k_clusters

NA

unpacking the tibble to have a look at the tot.withinss which is my key variable.

comp_clusterings <- k_clusters %>%
  unnest(glanced)

comp_clusterings

ggplot(comp_clusterings, aes(x=k, y=tot.withinss)) +
  geom_point() +
    geom_line() +
    scale_x_continuous(breaks = seq(1, 20, by = 1))

as i predicted….4. so the plot. below is as good as it gets.

computers_scale %>%
  kmeans.ani(centers = 4)

comp_clusterings %>% 
  unnest(cols = c(augmented)) %>%
  filter(k <= 4) %>%
 ggplot(aes(x = ram, y = hd)) +
  geom_point(aes(color = .cluster)) + 
  facet_wrap(~ k)

still unsure if clustering was suitable given the weird distribution. k = 3 might have been an option too.